2. Materials and methods

The studies were carried out in the years 1997, 1998 and 1999. The observations were made in early summer. The experimental groups of bees (Apis mellifera L.) were fed with the following components: bee pollen with sugar candy (natural proteinaceous food, total protein level = 20%), pollen substitute, and sugar candy only (non- proteinaceous food). The pollen substitute was a multicomponent food supplemented with methionine and lysine amino acids to reach the level found in bee pollen, and a total protein content of 20% similar to that of the bee pollen used. This pollen substitute was made of the following raw materials: potato protein (32%), extracted rape meal (6%), soya powder (18%), yeast Candida utilis (6%), wheat meal (14.8%), maize grits (17.5%). The basic components were supplemented with vitamin and mineral mixes Polfamix W (1.4%), vitamin preparation (0.3%), amino acids (methionine and lysine), and lactic acid to lower the pH of the pollen substitute to 5.5 as that of the bee pollen used. All components were ground to particles smaller than 150 µm. The energy source in the diet was soybean oil (3.5%) and rape lecithin (0.5%) as emulsifier. The fat fraction in the substitute was stabilized with an antioxidant (commercial name Rendox^®). Finally the pollen substitute was mixed with powdered sugar (1:1 w/w).

The total protein content in the diets was determined by the Kjeldahl’s method multiplying the nitrogen amount by 6.25. The amino acids were identified using an AAA339 analyzer (Mikrotechna, Czeck Republic) after a hydrolysis of the samples in 6 n HCl at 105 ºC for 23 h. Amino acids with sulphur were determined after their oxidation and fixation with performic acid. On the basis of amino acid composition, the index of each diet was calculated according to Chemical Score (Block and Mitchell, 1946) and Essential Amino acids index (Oser, 1952).

In each experimental group, one-day-old worker bees were fed for 7–8 days ad libitum in incubators at 28 °C and 40% RH. Young bees were kept without queens in small polystyrene cages (70 mm 80 mm 120 mm). Each group in each year consisted of 5 times 150 bees each, which originated from one queen and one colony. From each group, bees were randomly selected after 7– 8 days and their haemocytes were examined in haemolymph drawn from the sinus dorsalis of their circulatory system. Total haemocyte counts in undiluted haemolymph were determined using a Bürker’s haemocytometer. Every year, 7 to 21 observations of randomly selected bees were carried out for each diet; which gave a total of 37– 38 observations over 3 years.

Differential haemocyte counts were evaluated in smears of haemolymph stained with Wright’s stain according to the method of Gilliam and Shimanuki (1971). Following Jędruszuk (1997), haemocytes were classified as plasmatocytes, granular haemocytes, or other (all other types of haemocytes pooled together; Fig. 1). Plasmatocytes were small round cells with compact, round nuclei and thin, hyaline neutrophilic or pale-basophilic cytoplasm or in which the cytoplasm was not visible. Granular haemocytes were large oval or ellipsoidal cells with granular nuclei and vacuolated cytoplasm. Their cytoplasm was neutrophilic or pale-basophilic and occupied most of the cell. Other types of haemocytes were large (from several times to twelve or more times bigger than the plasmatocytes). They made less than 10% of the total number of haemocytes. The percentage of haemocytes of each type was calculated in each sample. From 8 to 36 smears of haemolymph were examined for each diet every year to give a total of 50 to 70 over three years.

Metabolic activity was observed in the haemocytes using a modified Nitro Blue Tetrazolium (NBT) test (Gliński and Klimont, 1987b). The method is based on the reduction of nitro blue tetrazolium to phormasane in haemocytes showing signs of metabolism indicating that they are active in phagocytosis. A drop of haemolymph was mixed with 2.0 L of the incubation solution

Influence of diet on honeybee haemocytes 99

Table I.I Cell content of haemolymph of worker honey bees raised in standard free-flying colonies and in incubator on different diets.

Bioassay with special diet (7–8 d old bees; summary of 1997–1999)

Group 5-d old workers in standard colonies

(1997)

Diet with bee pollen Diet with pollen

substitute

Diet with sugar only

Total haemocyte Count (per µL)

15172 + 4507

n = 50

15287 + 8984 a b^†

n = 38

14231 + 5039 b

n = 38

16574 + 4446a

n = 37

Plasmatocytes (%)

85.3 + 17 ( n = 48) 76.4 + 23.1 b (n = 50) 86.2 + 14.6 a (n = 70)

63.8 + 29.6c (n = 63)

Granular haemocytes (%)

Other haematocytes (%)

Metabolic activity

(% positive with nitro- blue tetrazolium test)

13.3 +16.8 (n = 48) 21.1 + 23.5 b (n = 50) 12.0 + 14.7 c (n = 70) 35.1 + 29.8 a (n = 63)

1.4 +1.4 (n = 48) 2.5 + 2.5 a (n = 50) _{1.8 + 2.3 a (n = 70)} 1.1 + 1.6 a (n = 63)

*3.5 + 1.8 (n = 24) 2.6 + 2.3 a (n = 96) 1.7 + 1.9 b (n = 101) 0.9 + 1.3 c (n = 109)

^† Means ± sd; values followed by different letter in each line are significantly different at P = 0.05 based on Fisher’s LSD procedure.

* indicates significant differences (P < 0.05) between the values of the bees from the standard colonies and those from bees fed on pollen substitute based on a Student’s t-test.

(0.2% of NBT dissolved in 0.85% NaCl and phosphorane buffer pH 7.4) on a microscope slide. Then the mixture was incubated in a water bath for 15 min at 36 °C. Afterwards, the mixture was incubated in a water bath at room temperature for 15 min. During incubation, the mixture was stirred a few times by delicately tilting the glass slide. After incubation, the mixture was smeared on the glass slide and desiccated. The smear was then stained in Wright’s stain and observed under 1000  microscope magnification. The cells containing blue grains of phormasane were counted and recorded as a percentage of the total number of haemocytes. From 11 to 70 examinations of NBT- test were carried out every year for each diet resulting in a total of 96 to 109 observations over 3 years.

The number of haemocytes were analyzed after a square root transformation and the proportion of granular haemocytes and plasmatocytes after an Arcsin transformation to stabilize the variance among groups (Bliss, 1937). The years, diets, the interaction between years and diets were taken into consideration in the analysis of variance. I assumed a mixed model with a random effect for year. When the F tests were significant (P < 0.05), means were separated by applying the Fischer’s least significant difference procedure. Untransformed mean values are given in Table I.

Additionally, in order to get some baseline data on the haemocyte system in worker bees, we used 5-day old workers originating from standard

colonies in 1997. To that end, several hundred emerging bees were marked and placed in standard honeybee colonies and, 5 days later, about 50 individuals were randomly selected from that group and the characteristics of their haemocyte system were determined using the same methods as above. The results were compared with those obtained from the workers fed with the pollen substitute using Student’s t-test.

3. RESULTS

   The characteristics of haemocytes in work- ers living in standard colonies in the field in 1997 are presented in Table I. In the haemol- ymph of bees fed with non-proteinaceous food, we observed a tendency for an increase in total haemocyte count, a significant increase in the percentage of granular haemocytes, a significantly lower amount in other types of haemocytes and a lower metabolic activity of haemocytes compared with the haemolymph of bees fed with pollen and pollen substitute (Tab. I).

In the haemolymph of bees fed with pollen substitute, there was a lower total haemocyte count, a significantly higher percentage in plasmatocytes, a significant decrease of gran- ular haemocytes, and a significant decrease of

100 B. Szymaś et al.